BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//wp-events-plugin.com//7.4.5//EN
TZID:America/New_York
X-WR-TIMEZONE:America/New_York
BEGIN:VEVENT
UID:961@fds.yale.edu
DTSTART;TZID=America/New_York:20261007T120000
DTEND;TZID=America/New_York:20261007T130000
DTSTAMP:20261004T172233Z
URL:https://fds.yale.edu/events/fds-colloquium-jason-klusowski-princeton/
SUMMARY:FDS Colloquium: Jason Klusowski (Princeton)\, "Statistical Attribut
 e Alignment for Black-Box Generative AI"
DESCRIPTION:\nAbstract: Generative AI offers a powerful way to create synth
 etic data for simulation\, training\, and evaluation. But what if a model 
 generates outputs with an attribute distribution that differs from what a 
 downstream task requires? For example\, medical research may require image
  collections with specified distributions of demographic groups or disease
  categories. This talk develops a statistical framework for attribute alig
 nment through output post-processing\, using only query access to the gene
 rative model. After observing the attributes of generated outputs\, we sel
 ect a batch whose attributes jointly follow a target distribution\, exactl
 y or approximately. We develop sampling algorithms and establish their asy
 mptotic optimality in expected model query cost as the desired batch size 
 grows.\n\n\n\n\n\nSpeaker Bio: Jason M. Klusowski&nbsp\;is an Associate Pr
 ofessor in the Department of Operations Research and Financial Engineering
  (ORFE) at Princeton University. He is also affiliated with the Princeton 
 Laboratory for Artificial Intelligence (AI Lab) and the Princeton Language
  and Intelligence (PLI) initiative. He studies the mathematical and statis
 tical foundations of machine learning and AI systems\, examining how data\
 , computation\, and model structure shape their capabilities and limitatio
 ns.\n\n\n\nBefore joining Princeton\, Jason was an Assistant Professor in 
 the Department of Statistics at Rutgers University–New Brunswick. He rec
 eived his Ph.D. in Statistics and Data Science from Yale University.\n\n\n
 \nJason serves on the editorial board of&nbsp\;Bernoulli\, the journal of 
 the Bernoulli Society. His research is partially supported by a Sloan Rese
 arch Fellowship in Mathematics and&nbsp\;NSF CAREER DMS-2239448\, and was 
 previously supported by&nbsp\;NSF DMS-2054808&nbsp\;and&nbsp\;TRIPODS DATA
 -INSPIRE Institute CCF-1934924.\n\n\n\nJason grew up in Winnipeg\, in the 
 heart of the Canadian Prairies. His&nbsp\;spouse&nbsp\;is an Assistant Pro
 fessor of Marketing at Yale University.\n
CATEGORIES:Fellows Events,FDS Events,Colloquium
LOCATION:Yale Institute for Foundations of Data Science & Webcast\, 219 Pro
 spect Street\, 13th Floor\, New Haven\, CT\, 06511\, United States
X-APPLE-STRUCTURED-LOCATION;VALUE=URI;X-ADDRESS=219 Prospect Street\, 13th 
 Floor\, New Haven\, CT\, 06511\, United States;X-APPLE-RADIUS=100;X-TITLE=
 Yale Institute for Foundations of Data Science & Webcast:geo:0,0
END:VEVENT
BEGIN:VTIMEZONE
TZID:America/New_York
X-LIC-LOCATION:America/New_York
BEGIN:DAYLIGHT
DTSTART:20260308T030000
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
END:DAYLIGHT
END:VTIMEZONE
END:VCALENDAR